HomeResources Videos Playmaker Video Detail Page

Paul Munding - SMART Gas Lift

Moneymakers Business Forum | 2019 Oklahoma City

Moneymakers Business Forum | 2019 Oklahoma City

Summary

SMART Gas Lift. A Moneymaker Forum panel discussion talk given by Paul Munding, Flogistix in Oklahoma City, Oklahoma on 4 April, 2019.

Full Transcript

Thanks, Susan. And thank all of you for being such diehards to stay here till the end.

Oh, OK. Oh, I see why now. All right. Anyways, I'm Paul Munding with Flogistix and I came here to join Flogistix about three years ago. And one of the first things my boss told me is we need to develop a smart gas lift reciprocating compressor system. I said, oh, boy, that's a little bit over my head.

But anyways, here we are three years later, and we've got a pretty good system in place. We've got a couple of field trials, actually, in the scoop and stack. I'm going to talk about those here in a second and try to get through this fairly quickly.

OK, so historically, if you look at where the attention has been focused in the industry-- and I've seen this in my own experience-- drilling and completion gets most of the attention, and rightly so with the amount of the AFE percent that they make up. I mean, typically, it's 80% of the AFE or 70% of the AFE is committed to drilling completions.

But if you look at production on these wells, especially the unconventionals, the LOE cost is pretty high. And if you look at the first 10 years of production LOE and discount that back to present value, even with inflation, it could be as much as 40% to 45% of lifetime cost of these wells. And it's really not being paid attention to like it should.

And if you look at the amount of stakeholders in the processes of drilling and completions, there's a lot of people paying attention to it. On production, not so much, especially when you got several hundred wells you're dealing with.

So you got to think about costs. How are costs defined? Well, yeah, if you're going to go out here and put equipment on your well to produce it, you've got to look at what that equipment costs. But at the same time, is that equipment running? I mean, we have a pretty low standard in the oil and gas industry for equipment, I'd say.

I kind of apologize here for these graphics. I didn't know they were going to be that out of sorts there. But for example, on air conditioning, when you buy, spend, or save $2,000 for an air conditioner that only ran 70% of time, I don't think anybody here would probably do that. Does anybody here own a car that's older than year 2000? All right. OK, you got one? Wow. We found one. All right. All right, Greg.

But that's how we're operating our oil and gas wells. A lot of these compressors we got here are from the '90s, right? They're only running 70% of the time. If you look at a well that's making 250 barrels of oil and at 500 MCF a day, at current commodities prices, one hour of downtime would cost the company $635 of revenue. One day would cost them 15,000 on that well. One week would cost them 100,000.

I got news for you. There's a big operator in the Permian Basin right now. It's averaging 70% runtime on their gas lift. That means if they have wells like this, they're losing $100,000 a month a well. So that's a lot of lost revenue for these operators.

Solution, smart artificial lift. Utilize smart equipment. I know we've had a lot of talks in here about big data, and machine learning, and analytics. And you can kind of get lost in all those terms. But utilize an autonomous device, a smart device that is connected to other devices or networks via wireless protocols.

And so what we have is we have an edge computer that we've put into our control panel. And you can program this. The domain expert, petroleum engineer, whoever, can put in their formulas they want to use. We use outflow correlations, Beggs and Brill, Hagedorn, and Brown, Gray. And we look at what are the bottom hole pressures of these wells that we're estimating versus the injection rates.

So utilizing on smart gas lift, utilizing smart edge computing with a smart systems compressor skid to respond to downhole conditions, right? Reservoir pressure's depleting. GORs are increasing over time, flow pass are closing. Sales line pressures, these are always fluctuating. I've never seen a well that had constant sales line pressure. Compressors are going down in the gathering station. Wells are being shut in for nearby fracks.

Variables are always changing, affecting, causing, slugging, and heading in these wells, leading to loaded well conditions. Really, a compressor needs to be able to respond to these dynamic conditions, rather than say, well, we're just going to set it and forget it and live with them, if you really want to optimize production.

Gas lift management historically, if you look at a classical approach, five years ago, when I was a production engineering manager at RKI, I'd be on the phone with a production superintendent and the pumper talking about, hey, we need about 600 MCF a day. It looks like that works pretty good over here on these wells. Let's take a look at how this well performs now. OK?

We make that change. Maybe a week later, we're looking at OK, let's make another change and see how that occurs. And then so that time to respond, you may have a pumper that's got 50 wells. He's got to get to all his wells before he can get to that particular well. And then you've got a production superintendent. He's got more than one well to look at. So it could be some time before you respond to each individual well's needs.

Versus a smart approach where you've got an engineer or production tech that's in there, is in there programming what the compressor could do on the PLC or in the edge device. And then that compressor can real-time watch the well performance and respond to what it needs.

So classical approach, again, you got an engineer on the phone, might be looking at his wells here, calling a pumper out on the field, saying, hey, I need you to go open the suction control valve a little bit. Well, unless this guy's sleeping out there in a sleeping bag, you're not going to have the ability to dynamically adjust this well 24/7, right? It just doesn't make sense anymore with all of this edge computing power that we have.

So the new approach is to have this integrated skid. You've got an edge computer device here in the PLC panel. You've got integrated suction control valve, integrated injection flow rate, and let the compressor control the situation. What you're trying to achieve here is the ideal injection rate.

This is a hydrostatic gradient, right? If you add too much injection gas, now you've increased the friction on the system. If you add too little, now the well's loading up due to the gravity. So you're trying to look for that saddle point where you get the lowest bottom hole pressure, yet you have the best injection rate.

This is actually a well in the scoop that we were controlling off of a bottom hole pressure sensor. And so as you can see here, the scale on the left is from 1,800 pounds down to 900 pounds. Bottom hole pressure came on at 1,300 pounds. It was falling off. We were using the injection rate on the compressor, looking at what the bottom hole pressure did as we decreased our injection rate over different periods of time. As long as that bottom hole pressure continued to decline, we kept decreasing the injection rate.

What we found here is we didn't need to be injecting 600 MCF a day on this well. Actually, the best was probably closer to 250, 300. That meant a difference of about 10 barrels a day in added production by taking that friction out of the system. So this was actually worth pretty good value that we didn't even realize.

Here's an example of a scoop well with surface conditions impacting it, OK? So what you have here is the left axis is injection rate from 0 to 1,400 MCF a day and line pressure from 0 to 1,400 PSI. On the right axis is oil production, 325 to 445 barrels a day. So here's your line pressure in your field impacting the well, on this particular well. And it's running normally about 150 pounds, but here, it spikes up to 250 and then falls down here to about 80 pounds.

What you see here is OK, let's say we're operating the old style way. We're just going to say well, we're going to set 550 MCF a day of injection and forget it, right? So this is this light green production is what you would expect to achieve with that flowing overhead pressure according to this line pressure. The gold curve here is what the optimized injection rate tells you need based on my formulas, OK? That, the dark green curve, would be the oil produced based on that optimized injection rate.

So if you look at it from a bar chart-- and here's the line pressure again. Here, it's scaled from 0 to 300 on the right axis, and here's your increase in oil and gas on the left axis from 0 to 100, so barrels per day and MCF a day. If you look at the bar chart on a day-to-day basis, these are the differences you're able to achieve by letting the smart gas lift compressor inject the rate it really needs to accommodate the line pressure differences.

So in closing, smart gas lift allows smart systems to manage day-to-day operations that require more detail. We don't have the amount of eyeballs that we need on the production side to keep up with all these little dynamic changes happening on these wells. This will allow maximum runtime and efficiency. And then use smart devices to get the most production out of your wells at the end of the day. That's it. Thanks.

So I have a question for Paul. So what kind of results-- how has this improved some of the efficiencies in your wells?

Yeah, I would say that one particular example we have looked at in the scoop, it's probably added about, with just operating efficiency, probably at 300 barrels of oil a month over the first couple months. And then on runtime alone, that's hard to say-- probably another several hundred.

That's really impressive. It's barrels of oil or barrels of fluid?

Barrels of oil. Yeah.

OK, so not just barrels of water.

Mhm. Correct.

That's good. And what kind of well was it? I mean, how much was the-- did it go from, like, 200 to 500 or 800?

So yeah, on the operating efficiencies, you're typically looking at about 5% increase. So whatever amounts you're running, if it's 300 barrels a day, then maybe 10 to 15 increase, somewhere in there.

Wow. So did it pay for-- how quick was the payoff for the actual installation of the equipment?

This one, I would estimate-- let's see-- it would be three to four days, maybe.

Wow, that's good.

Yeah.

So are both you and Tony interested in having companies contact you to see about developing intentional wells?

Mhm.

We would, for sure. Yeah.

OK, great.

Would you say how low the percentage is for the smart system as opposed to the standard system? How much more does it make?

Yeah, I mean, it's really going to vary depending on your field conditions. But typically, like I said, in the Permian, they only have a 70% runtime. So just right there, you could pick up another 20% to 30% increase, I believe. And then on any kind of optimization, another 5% to 10%.

Yeah, I was wondering after the cost.

Oh, after the cost.

Yeah.

It would be, oh-- like, for an unconventional well, that would probably knock off just a couple of percent off of, like, a 30% increase, so.

Oh, put the system on the well? If you're renting it, for the q, that could be $7,500. That's our bigger 200 horsepower unit. For the p, which is the one, I believe, 140, it's about 6,500 a month.

So how does that compare with the non-smart?

It's a couple of thousand more than the non-smart. But again, I mean, if your well's making $635 an hour and the non-smart's down for another 100 hours, then I don't think it really matters. The difference there's going to be negligible.

Show more

See Also ...

DPA Videos

Other Videos